Periodic error from Netty Epoll Acceptor IO Thread

[Bobrokus]

Describe the problem

I'm getting the same error periodically on Prominence II modpack server:

[Netty Epoll Acceptor IO Thread/ERROR]: Exception occurred in netty pipeline
io.netty.channel.unix.Errors$NativeIoException: recvAddress(..) failed: Connection reset by peer

Based on a response to other user's Reddit post the error is possibly caused by mc-monitor:

Apparently this is “itzg/mc-monitor” from the same developer, which is built into the image to monitor the server status. For some reason it doesn't work properly with the "Prominence" modpack.

A quick and temporary solution looks like this:

docker exec -it mc-prominence-container rm /usr/local/bin/mc-monitor

After this the error will disappear. In a good way, you need to understand how to disable this check on a permanent basis, because after updating the image the problem will return.

Workaround:

UPD. For a server based on itzg/docker-minecraft-server, just add the environment variable DISABLE_HEALTHCHECK: true.

Container definition

services:
  mc:
    image: itzg/minecraft-server:java17
    tty: true
    stdin_open: true
    ports:
      - 25565:25565
    environment:
      SERVER_HOST: localhost
      SERVER_POST: 25565
      EULA: TRUE
      TYPE: AUTO_CURSEFORGE
      MEMORY: 12G
      SPAWN_PROTECTION: 1
      ONLINE_MODE: false
      USE_AIKAR_FLAGS: true
      TZ: Europe/Bratislava
      DIFFICULTY: 2
      OPS: |
        Bobrokus
      MODRINTH_PROJECTS: |
        spark
        netherportalfix
        chunky
        notenoughcrashes
        memoryleakfix
        servercore
        krypton
        c2me-fabric:alpha
        vmp-fabric:alpha
        starlight
        noisium
        fastload
        dynmap
      CF_PAGE_URL: https://www.curseforge.com/minecraft/modpacks/prominence-2-rpg
      CF_API_KEY: *hidden*
      CF_EXCLUDE_MODS: |
        welcome-screen
        config-menus-forge
        night-config-fixes
        neruina
      ALLOW_FLIGHT: true
    volumes:
      - /root/mc:/data

Container logs

_minecraft-server-mc-1_logs.txt

[itzg]

I am going to convert to a discussion since I assume you were raising the Reddit post for awareness?

[itzg]

If you would like me to continue any investigation, then you need to provide the requested info in the template that you left blank.

[Bobrokus]

If you would like me to continue any investigation, then you need to provide the requested info in the template that you left blank.

Yea sorry I'm working on it right now

[Bobrokus]

If you would like me to continue any investigation, then you need to provide the requested info in the template that you left blank.

Done. Sorry for the delay I forgot to paste it right away.

[itzg]

I am not able to recreate the "Netty Epoll" error. Those logs don't even come from my image software -- it just runs the minecraft server after setting up mod contents.

Last logs I see are

[01:47:20] [Server thread/INFO]: Configured OPAC to use the following party system as primary: default
2025-07-21T01:47:20.819487168Z [01:47:20] [Server thread/INFO]: Registered synced player tracker system: ftb_teams
2025-07-21T01:47:20.824655648Z [01:47:20] [Server thread/INFO]: Registered synced player tracker system: ftb_teams
2025-07-21T01:47:20.832151533Z [01:47:20] [Server thread/INFO]: Auto report config loaded for mod: "copycats"
2025-07-21T01:47:20.832287821Z [01:47:20] [Server thread/INFO]: Auto report config loaded for mod: "the_bumblezone"
2025-07-21T01:47:20.836128594Z [01:47:20] [Server thread/INFO]: Blacklist loaded for namespace: "the_bumblezone"
2025-07-21T01:47:20.845200806Z [01:47:20] [Server thread/INFO]: Mixing common.MixinServerStatus from mixins/common/nochatreports.mixins.json into net.minecraft.class_2926
2025-07-21T01:47:20.856612619Z [01:47:20] [Server thread/INFO]: Loading Xaero's Minimap - Stage 2/2 (Server)
2025-07-21T01:47:20.856759535Z [01:47:20] [Server thread/INFO]: Loading Xaero's World Map - Stage 2/2 (Server)
2025-07-21T01:47:22.456123386Z [01:47:22] [Server thread/WARN]: Can't keep up! Is the server overloaded? Running 2038ms or 40 ticks behind
2025-07-21T01:47:29.547717544Z [01:47:29] [C2ME worker #3/WARN]: Non-existent template pool reference: ctov:village/bounty/desert
2025-07-21T01:47:29.551419040Z [01:47:29] [C2ME worker #3/WARN]: Non-existent template pool reference: ctov:village/vampirism/totem
2025-07-21T01:47:29.555130547Z [01:47:29] [C2ME worker #3/WARN]: Non-existent template pool reference: ctov:village/waystone/sand
2025-07-21T01:47:30.460321360Z [01:47:30] [Server thread/ERROR]: POI data mismatch: already registered at class_2338{x=-137, y=-31, z=689}

and container reports healthy:

NAME                    IMAGE                          COMMAND    SERVICE   CREATED         STATUS                   PORTS
prominence-2-rpg-mc-1   itzg/minecraft-server:java17   "/start"   mc        4 minutes ago   Up 4 minutes (healthy)   0.0.0.0:25565->25565/tcp, [::]:25565->25565/tcp

Since you are seeing those errors every 30 seconds it does seem likely that something isn't compatible with https://minecraft.wiki/w/Java_Edition_protocol/FAQ#What_does_the_normal_status_ping_sequence_look_like?

In your case, disable health check by setting the env var DISABLE_HEALTHCHECK to "true".

[Bobrokus]

Since you are seeing those errors every 30 seconds it does seem likely that something isn't compatible with https://minecraft.wiki/w/Java_Edition_protocol/FAQ#What_does_the_normal_status_ping_sequence_look_like?

Actually it's exactly every 1 minute.